Biophysical modeling as a translational bridge for understanding neural ensemble alterations in schizophrenia. - SUMMARY/ABSTRACT Schizophrenia is a devastating and burdensome illness that afflicts ~1% of the global population. Cognitive symptoms are a hallmark of the disease, affecting most individuals with schizophrenia, and being responsible for the greatest reduction in quality of life. Despite their significant impact, the biological mechanisms of cognitive deficits remain elusive, in part due to limitations of the experimental approaches typically used to study them in humans. To overcome these limitations, we propose a novel approach using biophysical modeling as an explanatory theoretical framework for bridging the translational gap between previous preclinical work in mouse models of schizophrenia-relevant risk and the proposed work in patients with schizophrenia. We propose translation of the findings of reduced neuronal ensemble reliability (n-ER) in the primary visual cortex (V1) as a window into a brain-wide circuit-level alteration in schizophrenia and its relationship to cognitive deficits. To achieve this, we will use a combined sample of 1,760 individuals, including healthy individuals, patients with schizophrenia or bipolar disorder and their first-degree relatives, from the HCP Young Adult, HCP Psychosis, and HCP Early Psychosis projects. Specifically, we will measure voxel ensemble reliability (v-ER) in humans using resting-state and visual-stimulation fMRI data—akin to calcium imaging studies in mice— as a theoretically grounded and translational index of excitation-inhibition balance (E/I) in cortical circuits. First, we aim to develop a biophysical model of V1 constrained by preclinical and basic neuroscience experiments, and test model predications of neuroimaging measures related to E/I. Second, we will test for reduced v-ER in patients with schizophrenia—directly translating preclinical findings—and use biophysical model simulations to identify potential biological mechanisms. Third, we will use the unique sample characteristics of the HCP Psychosis project (patients and first-degree relatives) to investigate the relationship between genetic burden for schizophrenia and v-ER. Fourth, given the convergence of cognitive deficits in the preclinical mouse models, we will examine the relationship between v-ER and cognitive performance. We will further seek to establish reduced v-ER as a brain-wide mechanism of cognitive deficits by testing for relationships in cognition across disparate sensory domains. Throughout, we will use well-powered, rigorous, state-of-the-art fMRI and statistical data-driven methods suitable for large-scale studies and HCP-like fMRI sequences, including cross-validation and independent confirmation. Together with a strong theoretical foundation and using biophysical modeling to complement fMRI analyses, this approach will begin to elucidate the biological mechanisms of cognitive deficits in schizophrenia. In doing so, this project will establish v-ER as a fully translational neuroimaging measure with the potential to be used as a biomarker for treatment selection and target engagement and will generate predictions that can be directly tested in preclinical studies.
